DOI QR코드

DOI QR Code

Methodology for Apartment Space Arrangement Based on Deep Reinforcement Learning

  • Cheng Yun Chi (Geospatial Analytics & Monitoring Center, Korea Research Institute for Human Settlements) ;
  • Se Won Lee (Geospatially Enabled Society Research Center, Korea Research Institute for Human Settlements)
  • Received : 2024.02.16
  • Accepted : 2024.03.19
  • Published : 2024.03.30

Abstract

This study introduces a deep reinforcement learning (DRL)-based methodology for optimizing apartment space arrangements, addressing the limitations of human capability in evaluating all potential spatial configurations. Leveraging computational power, the methodology facilitates the autonomous exploration and evaluation of innovative layout options, considering architectural principles, legal standards, and client re-quirements. Through comprehensive simulation tests across various apartment types, the research demonstrates the DRL approach's effec-tiveness in generating efficient spatial arrangements that align with current design trends and meet predefined performance objectives. The comparative analysis of AI-generated layouts with those designed by professionals validates the methodology's applicability and potential in enhancing architectural design practices by offering novel, optimized spatial configuration solutions.

Keywords

References

  1. AlOmani, A., & El-Rayes, K. (2020). Automated generation of optimal thematic architectural layouts using image processing. Automation in construction, 117, 103255. 
  2. As, I., Pal, S., & Basu, P. (2018). Artificial intelligence in architecture: Generating conceptual design via deep learning. International Journal of Architectural Computing, 16 (4): 306-327.  https://doi.org/10.1177/1478077118800982
  3. Choi, J., Cho, H., Choi, H., Park, I., & Funahashi, K. (2005). Spatial Configuration Analysis Using the Eigenvector Ratio of Adjacency Matrix (Summaries of the 10^ annual meeting). MERA Journal, 9(1): 36. 
  4. Combes, L. (1976). Packing rectangles into rectangular arrangements. Environment and Planning B: Planning and Design, 3 (1): 3-32.  https://doi.org/10.1068/b030003
  5. Flemming, U. (1978). Wall representations of rectangular dissections and their use in automated space allocation. Environment and Planning B: Planning and Design, 5 (2): 215-232.  https://doi.org/10.1068/b050215
  6. Grason, J. (2021). Fundamental description of a floor plan design program. In EDRA 1 (pp. 175-180). Routledge. 
  7. Gu, H. M., Seo, J. H., & Choo, S. Y. (2019). A Development of Facade Dataset Construction Technology Using Deep Learning-based Automatic Image Labeling. Journal of the architectural institute of Korea planning & design, 35 (12): 43-53. 
  8. Hamid-Lakzaeian, F. (2019). Structural-based point cloud segmentation of highly ornate building facades for computational modelling. Automation in Construction, 108, 102892. 
  9. Hillier, B., & Hanson, J. (1989). The social logic of space. Cambridge university press. 
  10. Homayouni, H. (2007). A genetic algorithm approach to space layout planning optimization (Doctoral dissertation, University of Washington). 
  11. Huang, W., & Zheng, H. (2018, October). Architectural drawings recognition and generation through machine learning. In Proceedings of the 38th annual conference of the association for computer aided design in architecture, Mexico City, Mexico (pp. 18-20). 
  12. Jin, C., Xu, M., Lin, L., & Zhou, X. (2018). Exploring BIM Data by Graph-based Unsupervised Learning. In ICPRAM (pp. 582-589). 
  13. Jo, J. H., & Gero, J. S. (1998). Space layout planning using an evolutionary approach. Artificial intelligence in Engineering, 12 (3): 149-162.  https://doi.org/10.1016/S0954-1810(97)00037-X
  14. Jung, K., & Kim, I. (2021). C-COMA: A Continual Reinforcement Learning Model for Dynamic Multiagent Environments. KIPS Transactions on Software and Data Engineering, 10 (4): 143-152. 
  15. Kim, C. H., & Park, T. Y. (1995). A theoretical study on the principles of Computer-Aided Space Planning technique. JOURNAL-ARCHITECTURAL INSTITUTE OF KOREA, 11: 55-70. 
  16. Kim, Y. S. (2021). Control of Lighting System based on Illuminance Prediction Model with Deep Deterministic Policy Gradient (DDPG), Master's Degree Thesis, Seoul National University 
  17. Kim, J. S., & Lee, J. K. (2020). Implementation and application of interior design style training model using deep learning. Journal of the Korean institute of interior design, 29 (5): 96-104.  https://doi.org/10.14774/JKIID.2020.29.5.096
  18. Kim, J., & Lee, J. (2019). Implementation of auto-classification of unclassified objects in BIM model using 2D CNN for design rule-checking systems. Korean Journal of Computational Design and Engineering, 24 (4): 452-461.  https://doi.org/10.7315/CDE.2019.452
  19. Kim, H. J., Ji, S. Y., & Jun, H. (2019). A Study on Application of Artificial Intelligence Technology to BIM Architectural Planning-Focus on Structural BIM Model in Early Design Phase. Korea Soc. Art Des, 22, 229-242. 
  20. Kim, Y. G., Heo, K., You, G. E., Lim, H. S., Choi, J. I., Ku, K. D.,... & Jeon, Y. S. (2018). A Study on the Improvement of Heat Energy Efficiency for Utilities of Heat Consumer Plants based on Reinforcement Learning. Journal of Energy Engineering, 27 (2): 26-31.  https://doi.org/10.5855/ENERGY.2018.27.2.026
  21. Kim, J., & Lee, H. (2016). Multi-agent Reinforcement Learning based Evacuation Framework Considering Both Evacuation Time and Crowdedness. J. Korean Inst. Intell. Syst, 26, 335-342. 
  22. Kim, I. H., Yang, J. I., Cho, G. H., & Choi, J. S. (2012). A study on the BIM-based Automation of Envelope Form Generation at the Schematic Design Phase of Super-tall Buildings. Journal of the Architectural Institute of Korea Planning & Design, 28 (11): 11-18. 
  23. Kwon, O., & Cho, J. (2019). Space Usage Knowledge Extraction from BIM Data by Decision Tree and Expert System. Korean Journal of Computational Design and Engineering, 24 (2): 126-134.  https://doi.org/10.7315/CDE.2019.126
  24. Kwon, O., & Cho, J. (2020). Quantitative Comparison of BIM Architectural Space Designs by Decision Tree and Expert System. Korean Journal of Computational Design and Engineering, 25 (1): 36-44.  https://doi.org/10.7315/CDE.2020.036
  25. Lee, S. H., & Chi, C. Y. (2021). AI-based Spatial Arrangement Simulator with Reinforccement Learning. Journal of the Architectural Institute of Korea, 37 (11): 43-53. 
  26. Lee, S. H., & Lu, N. (2020). A Methodology of Enhancing the Accuracy of Image Classification with CNN. Journal of the Architectural Institute of Korea, 36 (9): 15-22. 
  27. Lee, Y. G. (2012). A basic study on the development of the technology of simulating the user's behavior in the real time manner within the BIM-based design process. Journal of the Architectural Institute of Korea Planning & Design, 28 (12): 157-164. 
  28. Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., Veness, J., Bellemare, M. G., ... & Hassabis, D. (2015). Human-level control through deep reinforcement learning. nature, 518 (7540): 529-533.  https://doi.org/10.1038/nature14236
  29. Nguyen, K., Le, M., Martin, B., Cil, I., & Fookes, C. (2022). When AI meets store layout design: a review. Artificial Intelligence Review, 55 (7): 5707-5729. 
  30. Rahbar, M., Mahdavinejad, M., Bemanian, M., Davaie Markazi, A. H., & Hovestadt, L. (2019). Generating synthetic space allocation probability layouts based on trained conditional-GANs. Applied Artificial Intelligence, 33 (8): 689-705.  https://doi.org/10.1080/08839514.2019.1592919
  31. Sun, Y. (2022). Design and optimization of indoor space layout based on deep learning. Mobile Information Systems, 2022, 1-7.  https://doi.org/10.1155/2022/2114884
  32. Schiller, E. (2018). Creating Novel Architectural Layouts With Generative Adversarial Networks (Doctoral dissertation, Harvard University). 
  33. Solihin, W., & Eastman, C. (2015). Classification of rules for automated BIM rule checking development. Automation in construction, 53, 69-82.  https://doi.org/10.1016/j.autcon.2015.03.003
  34. Schmidhuber, J. (2015). Deep learning in neural networks: An overview. Neural networks, 61: 85-117.  https://doi.org/10.1016/j.neunet.2014.09.003
  35. Turner, A. (2001, May). Depthmap: a program to perform visibility graph analysis. In Proceedings of the 3rd international symposium on space syntax (Vol. 31, pp. 31-12). Atlanta, GA, USA: Georgia Institute of Technology. 
  36. Upasani, N., Shekhawat, K., & Sachdeva, G. (2020). Automated generation of dimensioned rectangular floorplans. Automation in Construction, 113, 103149. 
  37. Wan, D., Zhao, X., Lu, W., Li, P., Shi, X., & Fukuda, H. (2022). A Deep Learning Approach toward Energy-Effective Residential Building Floor Plan Generation. Sustainability, 14 (13): 8074. 
  38. Wan, T., & Ma, Y. (2022). Urban Planning and Design Layout Generation Based on Artificial Intelligence. Mathematical Problems in Engineering, 2022. 
  39. Yang, F., Li, L., Su, F., Li, D., Zhu, H., Ying, S., ... & Tang, L. (2019). Semantic decomposition and recognition of indoor spaces with structural constraints for 3D indoor modelling. Automation in Construction, 106, 102913. 
  40. Yoshimura, Y., Cai, B., Wang, Z., & Ratti, C. (2019). Deep learning architect: Classification for architectural design through the eye of artificial intelligence. Computational Urban Planning and Management for Smart Cities 16, 249-265.  https://doi.org/10.1007/978-3-030-19424-6_14